Re: [bug-gnulib] <stdint.h> on IRIX-5.3 & SGI-cc

[Bruno Haible]

Martin Neitzel wrote:
> This is CVS on a SGI Indigo (R3000, 32bit) running IRIX-5.3 and SGI's
> commercial "IRIX Developer Option" cc  (ANSI C as in ISO :1990 or
> c89).

OK, we can support this platform.

> I'm using -w2 which turns all warnings into serious errors
> requiring some treatment (just like gcc -Werror does).

But this is something we cannot support. Various compilers give
hardly justifiable warnings here and there. We try to minimize the
warning on those platforms on which most developers work. But it
would take too much effort to make gnulib warning-free on all platforms.

For a platform such as IRIX 5, we can attempt to make gnulib code compile
fine and run fine, but asking for warning-free-ness is too much.

> On this platform, the following SGI system headers exist:
> <sys/types.h>:        typedefs int{8,16,32}_t
> <inttypes.h>: typedefs int{8,16,32,64}_t & unsigned counterparts
> 
> These two headerfiles are mutually exclusive.  The comments in config.h
> indicate this possible clash, and the autoconfiguration #defines neither
> HAVE_INTTYPES_H nor HAVE_INTTYPES_H_WITH_UINTMAX (correctly, according
> to the comments).

OK.

> My adjustments to get md5.c to compile were these:
> 
> (a)  make the (generated) lib/stdint.h supply the missing "int64_t" and
>      all uintXX_t typedefs, but rely on int{8,16,32}_t coming from
>      SGI's <sys/types.h>.

The current rewritten stdint module should do fine out of the box now.

> (b)  reorder the includes in md5.c so that SGI's <sys/types.h>
>      is included before md5.h indirectly includes the (generated)
>      gnulib-stdint.h (the latter really depends on <sys/types.h>).

Likewise: the current rewritten stdint.h includes <sys/types.h> on IRIX.

Bruno